Saint-Gobain, a global leader in sustainable building materials and solutions, operates in over 70 countries with more than 170,000 employees. With a legacy spanning over 350 years, the company is now firmly focused on building a data-driven future – leveraging AI, advanced analytics, and modern data platforms to drive efficiency, innovation, and customer value across its vast industrial ecosystem.

In this second installment of our three-part series, Benoit Lepetit, Group Chief Data and Analytics Officer at Saint-Gobain, sits down with Julian Schirmer, Co-Founder of OAO, to explore how the organization is defining, scaling, and operationalizing data products across a decentralized, global enterprise. From the evolution of digital twins in manufacturing to harmonizing datasets across regions, Lepetit outlines Saint-Gobain’s pragmatic approach to building scalable, high-quality data assets that deliver measurable business value.

The conversation touches on Saint-Gobain’s federated data governance model, the cultural and organizational shifts required to embed data into the company’s DNA, and the real-world challenges of aligning local agility with global standards. At the heart of it all is a sharp focus on turning information into actionable, productized assets that serve both internal users and end customers – whether it’s a factory operator, a craftsman, or a business leader.

Edited Excerpts

Q: What do data products mean within the Saint-Gobain universe? Can you break it down and help us understand how it translates into action?

This concept of data products can be a bit difficult because there are different visions and levels of maturity. We define a data product as a set of capabilities built around a domain or specific aspect, designed to ensure a certain level of quality. The goal is for the consumer to be able to follow the full evolution of the information. The system needs to guarantee a level of quality that meets the demand so that the use of these products has a bigger impact.

Let me give an example. For us, a data product is a combination of datasets, the consumables built on top of those datasets (such as dashboards), and also the AI capabilities that are linked to a particular user need. It’s this combination that forms one product aimed at one consumer.

This ideal version of a data product is still maturing. Sometimes, we have fewer features or simpler products, starting from data capture to the dataset itself to modernization, and finally to the consumables. But all of it should be guided, assured, and built to maintain quality throughout the product’s lifecycle.

Q: Your data products and information capital serve both internal and external purposes. What implications does this dual focus have for your organization?

In the building industry, the craftsmen – for instance, those who are part of our value chain – are maybe less technologically savvy. So, we have a role as a leader in construction materials to support them in developing themselves. When we build solutions that facilitate the billing process for craftsmen toward their end users, it reflects our responsibility as an industry leader to help drive this transformation. We do this by bridging the technology gap and creating assets that our customers can use to accelerate their ability to adopt our products or solutions, in alignment with the group’s direction.

Q: Could you give us an example of a critical data product?

For example, digital twins are something we had for years in our flat glass businesses. It was not called digital twin then, but it was the same way of using sensor data, using ontology to classify and modernize this data, and having consumables like analytics and predictive analytics on top of that.

Such data products are fundamental for industrial operational efficiency – not only to accelerate transformation but also to reduce costs, accelerate production, enable agility towards our customers, build new products, reduce waste, and also enable circularity.

It was not called a digital twin before, but this concept has now been packaged as a data product that can be scalable in multiple domains – in multiple ways in the value chain of industrial production through our multiple lines of business.

Q: How do you manage to harmonize your data products across the organization?

It’s always the same – technology, process, people.

Speaking of technology, we need a modern architecture that allows us to leverage the scaling effect in how we use our data products. A cloud data platform is essential to scale this, and that’s what we’ve delivered: a data platform factory that can be replicated globally to cover all the regions of Saint-Gobain.

The second aspect is about methodology – processes where we need clear definitions with joint methodologies. But on top of that, we must focus on collaboration between data teams and business teams.

Last but not least, all of this operates under a data governance framework. This defines roles and responsibilities and also outlines the processes to identify what data ownership means within the business. It helps us determine who the data owners are and how they manage data products to ensure their relevance to operations. This framework guarantees quality, accessibility, compliance, and security, which allows us to scale across geographies where regulations may differ.

As a very decentralized group, we follow a federated data governance model. It gives us local agility while enabling convergence in applications. It also ensures we can scale securely, with the confidence that our data usage aligns with the compliance and security standards we need to protect the group from risk, while also delivering a positive impact and clear business value.

Q: What would you say are the key challenges in making this a reality?

The key challenges include linking data to business value. The first is to be obsessed with business value – to ensure that data products deliver measurable outcomes and also align with our strategic goals.

One common problem in any organization, especially in a highly decentralized one, is silos. We need to break down these organizational silos to create seamless data integration across business units, enabling us to leverage the strengths of our entire group.

There’s also the challenge of change management. We need to drive a cultural shift to promote data-driven decision-making among employees. We are an industrial group, and historically, data was not at the core of Saint-Gobain’s DNA.

Then, there’s the competition for talent. We need to address the talent gap – recruiting and retaining skilled data professionals who can help drive this critical innovation and transformation for our group.

CDO Magazine appreciates Benoit Lepetit for sharing his insights with our global community.